Abstract Bilateral loss of vestibular function (inner ear balance sensation) due to ototoxic hair cell injury is disabling, with patients suffering disequilibrium and inability to maintain stable vision during head movements typical of daily life. While most individuals with partial loss compensate through rehabilitative strategies enlisting other senses, those who fail to compensate for profound loss have no good therapeutic options. Because the vestibular nerve should be intact in many of these patients, electrical stimuli encoding head rotation should be able to drive the nerve and restore sensation of head movement, much like a cochlear implant restores auditory function. The proposed research is guided by two broad goals. The first is to advance development toward an implantable neuroelectronic prosthesis that restores function to people disabled by bilateral loss of vestibular sensation. The second is to drive the field of vestibular neurophysiology though increased understanding of how vestibular nerve activity encodes head motion and through development of technologies that enable use of previously impossible experimental paradigms. This project builds upon significant progress we have already made toward this goal, including: (1) development of a multi-channel, head-mounted prosthesis able to encode three-dimensional (3D) head rotation as electrical stimulation of three or more vestibular nerve branches;(2) characterization of the 3D angular vestibulo-ocular reflex (AVOR), vestibular nerve activity and endorgan histology in chinchillas after vestibular ototoxic injury via gentamicin treatment;and (3) partial restoration of the AVOR via prosthetic stimulation. These studies have identified channel interaction causing misalignment of eye and head rotation as a key challenge to restoration of a normal 3D aVOR. We hypothesize that misalignment is mainly due to spurious electrical stimulation of bystander vestibular nerve branches by inadequately selective electrodes. In this project, we will: (1) characterize the dependence of 3D AVOR eye rotations on stimulus parameters;(2) determine the extent and time course of adaptation to chronic prosthetic input;(3) develop and validate a precise model of current flow paths in the implanted labyrinth;and (4) extend our studies from chinchillas to macaque monkeys, which have inner ear dimensions similar to humans. We hypothesize that implanted macaques will exhibit much less misalignment than do chinchillas, and that the modeling and design techniques developed in chinchillas can generalize accurately to primates. Through extrapolation of electrode designs, stimulus optimization protocols, finite element models and surgical techniques from rodents to nonhuman primates, this project will set the stage for rational design and initial clinical studies of a multichannel vestibular prosthesis to aid individuals disabled by loss of vestibular sensation.